Africa: the continent the rest of the world’s $\mathcal{R}$ depends on

Africa is the substrate question of the 21st century. By 2050 it holds ~25 % of humanity; by 2100, ~38 %. It currently harvests ~4 % of the world’s primary energy. The gap is the central geopolitical fact of the next eighty years — for Africa’s own persistence and for everyone else’s, because the world’s demographic, agricultural, and migration $\Phi$ now flows through this continent’s accounting.

This essay treats Africa as a graph of 54 nodes, not a node. Generalisation is dangerous. The aggregate numbers are still informative.

What it is made of ($\Phi$)

Substrate	Magnitude	$\Phi$ status
Population	1.44 B (2024), 2.5 B by 2050, ~3.9 B by 2100 (UN medium)	Youngest population on Earth; median age ~19
Land	30.4 M km²	~20 % arable, much underexploited
Freshwater	Huge basins (Congo, Niger, Nile, Zambezi)	Distribution skewed; northern Africa stressed
Energy reserves	Major oil (Nigeria, Libya, Algeria, Angola, Mozambique gas), uranium (Niger, Namibia), solar potential unmatched on Earth	Largely under-developed at home; exported as raw
Critical minerals	Cobalt (DRC ~70 % world), platinum-group (SA ~70 %), manganese, lithium (Zimbabwe, DRC), copper (Zambia, DRC), graphite, REEs	Single largest mineral substrate on Earth, mostly mined by foreign capital
Manufacturing	~2 % of world value-add	Tiny in absolute terms; Egypt, Morocco, South Africa lead
Infrastructure	~600 GW total electrical capacity for 1.4 B people	Compare: USA has ~1.2 TW for 335 M
Universities	~3 in world top 500 (all South African)	Severe brain drain; ~70k African doctors abroad
Currencies	41 active national currencies, ~$1.5 T combined GDP	Highly fragmented; CFA franc + rand are mid-tier
Institutional shell	African Union, ECOWAS, SADC, EAC	Real institutions, modest $\Psi$ provision

The substrate is unprecedented in resource potential, structurally fragmented, and under-capitalised. The deepest constraint is not minerals or land or solar; it is functioning grids, ports, courts, and labour markets.

What it harvests ($P_{in} \cdot \eta$)

Primary energy: ~0.8 TW (~25 EJ/yr). About 4 % of world energy for 18 % of world population.
Per-capita primary power: ~570 W — fifteen times less than the USA, seven times less than the EU.
GDP (nominal): ~$3 T — about the size of France for a population four times larger.
$\eta$ — low and bimodal. South Africa, Egypt, Morocco are at middle-income $\eta$; the bulk of sub-Saharan Africa is in low-$\eta$ traps.
Cooking fuel for ~900 M people is biomass. This is the single largest unrecognised energy bottleneck: a population the size of Europe + USA combined burns wood and charcoal for daily heat, with the predictable $\Phi$-cost (indoor air, deforestation, women’s time).

A useful reframe: Africa’s $P_{in}$ deficit is the largest single accounting fact of the human population. The continent does not produce enough energy to host modernity for its own people, let alone its 2050 numbers.

Bureaucratic complexity ($\omega$) and friction ($\Gamma$)

$\omega$ varies enormously. Some states are sophisticated, capable bureaucracies (Rwanda, Botswana, Morocco, Mauritius, Senegal at moments). Others are skeletal. The cross-state average is insufficient institutional density for the size of the substrate — too few civil servants per square kilometre, too few engineers per population, too few judges per dispute.

$\Gamma$ items, in approximate order of regional weight:

Sahel collapse — Mali, Burkina, Niger now in coup/military regimes; Chad fragile; jihadi pressure rising; France pushed out; Russia / Wagner-successor entering with quasi-extractive deals.
Sudan civil war — ~30+ M displaced or food-insecure as of 2024–25; the largest current humanitarian collapse.
DRC east — ~30-year low-grade war over minerals.
Ethiopia / Tigray aftermath — ceasefire fragile; Amhara, Oromia simmering.
North Africa migration vector — Libya, Tunisia, Morocco function as queue points; politicises everywhere upstream.
Nigeria — biggest economy, biggest population, biggest fertility tailwind, biggest unsolved insurgencies (Boko Haram / ISWAP / Niger Delta / banditry).
South Africa — power crisis, infrastructure decay, governance corruption, capital flight; the continent’s most institutionalised state is in measurable decline.
Land + water — Nile dispute (Egypt / Sudan / Ethiopia GERD), Lake Chad shrinkage, Niger river basin, southern African drought cycles.
Debt — ~30 African states under IMF programmes or near-distress; Chinese, Western, and bondholder claims competing for first-loss seat.

What they believe about themselves ($\mathcal{D}_{KL}$)

Africa-as-graph has a smaller average $\mathcal{D}_{KL}$ than the great powers because most African governments have less margin for delusion — the consequences of misreading reality are felt fast and locally. The notable $\mathcal{D}_{KL}$ terms are at the continental level:

“African Century” narrative — partly justified by demographics, partly substitutes hope for industrial-policy execution. The demographic dividend doesn’t pay if the labour force has no factories to enter; the experience of much of sub-Saharan Africa 1990–2020 is jobless growth.
“China as the alternative to the West” — partially true, partially a $\mathcal{D}_{KL}$ on what Chinese capital actually wants (resources, ports, mines, repayment) versus what it is sold as (“development partner”). Each new Belt-and-Road default is a model-correction event.
AfCFTA — the African Continental Free Trade Area, ratified 2019, is the right idea. Intra-African trade is ~15 % of African trade; in the EU it’s ~60 %. Implementation has so far been mostly performative. Treating signed-but-not-implemented agreements as if they were implemented is a regional $\mathcal{D}_{KL}$ pattern.
State-level kleptocratic accounting — extracted rents counted as government revenue without modelling the elite-emigration of those rents to London / Dubai / Geneva. The hard accounting says: the substrate paid for the rent and didn’t get the road.
Climate-victim framing — true (Africa contributes ~3 % of emissions, suffers disproportionately) but politically used to delay the energy build-out that the continent itself needs. The framing is correct and paralysing.

Western and Chinese delusion about Africa is, on average, larger than African delusion about itself. The continent is routinely modelled as either tragedy-shaped or opportunity-shaped, rarely as the 54-node graph it is.

What shelters them ($\Psi$)

The remaining global trade order — commodity exports priced in USD, shipped on global lanes.
Remittance flows (~$100 B/yr to the continent, more than ODA + FDI in some years).
Diaspora networks — labour mobility to Europe, the Gulf, North America.
Foreign aid + WFP + WHO + UN agencies — directly addresses $\mathcal{E}_\Sigma$ shocks (famines, epidemics).
No African node is currently sheltered by another great power’s military umbrella — sovereignty is real, even if economic dependence is high.

The $\Psi$ on offer to African states from the Sino-Russian alternative (“non-interference,” cheap loans, security contracts) is being chosen by a growing fraction of governments because the Western $\Psi$ has visibly weakened (post-2003 Iraq legitimacy crash, French retreat from the Sahel, US withdrawal of attention). Whether this proves to be net $\Psi$ or net extraction is the open question of the 2030s.

What they actually keep doing well

Demographic vitality — every other major region is below replacement TFR; Africa is producing the world’s future workforce.
Mobile-payment leapfrogging — M-Pesa and successors made financial inclusion at $0 fixed cost work years before Western fintech.
Solar microgrids and pay-as-you-go solar — leapfrogging the grid where the grid never came.
Ethiopia, Rwanda, Côte d’Ivoire, Senegal, Morocco, Egypt have all sustained multi-decade ~5–7 %/yr growth periods.
Surviving — the continent has weathered colonial extraction, Cold War proxy wars, structural-adjustment austerity, multiple debt crises, the AIDS pandemic, and is still demographically and culturally robust. That is a high-$\mathcal{R}$ outcome.

Collapse trajectory if $\mathcal{R}$ slips below 1

Africa-as-aggregate does not collapse; it cannot. The trajectory question is node-by-node:

Failed-state ring — Sahel, parts of Central Africa, Sudan, Somalia, eastern DRC. $\mathcal{R} < 1$ already for several. They are sheltered by humanitarian $\Psi$ rather than restored to viability.
Climate-stressed coastal megacities — Lagos, Dakar, Alexandria — face combined sea-level + heat + freshwater stress; the population concentration there is the highest existential-$\mathcal{R}$ challenge to specific nodes.
Successful industrialisers — Morocco, Egypt, Ethiopia (post-civil-war), Rwanda, possibly Tanzania, Kenya, Côte d’Ivoire — can plausibly reach upper-middle-income status by 2050.

The aggregate trajectory is governed by one variable: whether the continent can build out electrical $P_{in}$ at ~10×–20× its current level over thirty years. With it, the demographic dividend converts into an industrial node. Without it, the dividend converts into an emigration node, which is the second-most-likely scenario and politically destabilising for everyone north of the Sahara.

The lowest-$\mathcal{D}_{KL}$ reform: treat electricity as the binding constraint and build it without ideological filter — gas, solar, hydro, nuclear, geothermal, all of it, at scale, with the institutional density that does maintenance.

Africa: the continent the rest of the world’s \(\mathcal{R}\) depends on

What it is made of (\(\Phi\))

What it harvests (\(P_{in} \cdot \eta\))

Bureaucratic complexity (\(\omega\)) and friction (\(\Gamma\))

What they believe about themselves (\(\mathcal{D}_{KL}\))

What shelters them (\(\Psi\))

What they actually keep doing well

Collapse trajectory if \(\mathcal{R}\) slips below 1

See also